Prior-exposure-test Experiment (all data combined)

Catch trial performance

## [1] "Excluded 1 participants based on catch-trial performance."

Load data from Exp 1.

trials = remove_quotes(format_data(read.csv("../../1_talker_specific_adaptation_prior_exp_test/data/1_talker_specific_adaptation_prior_exp_test-trials.csv")))
exp_trials = remove_quotes(read.csv("../../1_talker_specific_adaptation_prior_exp_test/data/1_talker_specific_adaptation_prior_exp_test-exp_trials.csv"))

conditions = read.csv("../../1_talker_specific_adaptation_prior_exp_test/data/1_talker_specific_adaptation_prior_exp_test-condition.csv")

subj_info = read.csv("../../1_talker_specific_adaptation_prior_exp_test/data/1_talker_specific_adaptation_prior_exp_test-subject_information.csv") %>% select(workerid, noticed_manipulation)

trials = merge(trials, conditions %>% select(workerid, condition), by=c("workerid"))

trials = merge(trials, subj_info, by=c("workerid"))

trials[, c("test_order", "first_speaker_type", "second_speaker_type", "confident_speaker")] = str_split(trials$condition, "_", simplify=T)



EXCLUDE_BELOW = 0.75

catch_trial_perf = get_correct_catch_trial_counts(exp_trials)



exclude = catch_trial_perf %>%
  filter(catch_prop < EXCLUDE_BELOW) %>%
  .$workerid


print(paste("Excluded", length(exclude), "participants based on catch-trial performance in Exp 1.."))

## [1] "Excluded 1 participants based on catch-trial performance in Exp 1.."

#final data
d.exp1 = trials %>% filter(., !(workerid %in% exclude))

#combine data from exp1 and exp3

d = rbind(d.exp1, d.exp3)

Aggregated results

## `summarise()` has grouped output by 'workerid', 'percentage_blue', 'modal'. You
## can override using the `.groups` argument.
## `summarise()` has grouped output by 'percentage_blue', 'modal'. You can
## override using the `.groups` argument.
## `summarise()` has grouped output by 'workerid', 'percentage_blue', 'modal'. You
## can override using the `.groups` argument.

Comparison across conditions

## `summarise()` has grouped output by 'workerid', 'percentage_blue', 'modal'. You
## can override using the `.groups` argument.
## `summarise()` has grouped output by 'percentage_blue', 'modal'. You can
## override using the `.groups` argument.

Individual responses

AUC computation

We use the AUC function with the splines method to directly compute the AUC.

t-test and regression model with control variables:

## 
##  Two Sample t-test
## 
## data:  aucs.cautious$auc_diff and aucs.confident$auc_diff
## t = 4.7434, df = 250, p-value = 3.536e-06
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##   9.08335 21.98205
## sample estimates:
## mean of x mean of y 
## 20.942809  5.410111

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: 
## auc_diff ~ cond + test_order + first_speaker_type + confident_speaker +  
##     first_speaker_type * cond + (1 | workerid)
##    Data: auc_d
## 
## REML criterion at convergence: 2301.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.99577 -0.46121  0.01379  0.57936  1.89279 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  workerid (Intercept) 194.3    13.94   
##  Residual             430.6    20.75   
## Number of obs: 252, groups:  workerid, 126
## 
## Fixed effects:
##                           Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)                13.4314     1.8051 122.0000   7.441 1.55e-11 ***
## cond1                       7.7724     1.3073 124.0000   5.945 2.61e-08 ***
## test_order1                 2.2133     1.8042 122.0000   1.227 0.222269    
## first_speaker_type1        -7.0973     1.8042 122.0000  -3.934 0.000139 ***
## confident_speaker1         -2.2698     1.8051 122.0000  -1.257 0.210990    
## cond1:first_speaker_type1  -0.3787     1.3073 124.0000  -0.290 0.772559    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cond1  tst_r1 frs__1 cnfd_1
## cond1        0.000                            
## test_order1  0.031  0.000                     
## frst_spkr_1 -0.015  0.000 -0.001              
## cnfdnt_spk1 -0.031  0.000  0.015 -0.031       
## cnd1:frs__1  0.000 -0.016  0.000  0.000  0.000

Clustering analyses

library(mclust)

## Package 'mclust' version 5.4.10
## Type 'citation("mclust")' for citing this R package in publications.

## 
## Attaching package: 'mclust'

## The following object is masked from 'package:DescTools':
## 
##     BrierScore

## The following object is masked from 'package:bootstrap':
## 
##     diabetes

aucs_diff = merge(aucs.cautious, aucs.confident, by=c("workerid"))
aucs_diff$diff_of_diffs = aucs_diff$auc_diff.x - aucs_diff$auc_diff.y

aucs_diff %>% ggplot(aes(x=diff_of_diffs)) + geom_density() + geom_jitter(aes(y=0), width=0, height=0.001)  + ggtitle("Raw data + estimated density")

Gaussian mixture models of diffeences of AUC differences

1 Cluster

fit1 = Mclust(aucs_diff$diff_of_diffs, G=1)
print(summary(fit1, parameters=2))

## ---------------------------------------------------- 
## Gaussian finite mixture model fitted by EM algorithm 
## ---------------------------------------------------- 
## 
## Mclust X (univariate normal) model with 1 component: 
## 
##  log-likelihood   n df       BIC       ICL
##       -603.5952 126  2 -1216.863 -1216.863
## 
## Clustering table:
##   1 
## 126 
## 
## Mixing probabilities:
## 1 
## 1 
## 
## Means:
## [1] 15.5327
## 
## Variances:
## [1] 848.101

2 Clusters

fit2 = Mclust(aucs_diff$diff_of_diffs, G=2)
print(summary(fit2, parameters=T))

## ---------------------------------------------------- 
## Gaussian finite mixture model fitted by EM algorithm 
## ---------------------------------------------------- 
## 
## Mclust E (univariate, equal variance) model with 2 components: 
## 
##  log-likelihood   n df       BIC       ICL
##       -593.5919 126  4 -1206.529 -1219.767
## 
## Clustering table:
##   1   2 
## 110  16 
## 
## Mixing probabilities:
##         1         2 
## 0.8586232 0.1413768 
## 
## Means:
##         1         2 
##  6.924947 67.810127 
## 
## Variances:
##      1      2 
## 398.11 398.11

3 Clusters

fit3 = Mclust(aucs_diff$diff_of_diffs, G=3)
print(summary(fit3, parameters=T))

## ---------------------------------------------------- 
## Gaussian finite mixture model fitted by EM algorithm 
## ---------------------------------------------------- 
## 
## Mclust E (univariate, equal variance) model with 3 components: 
## 
##  log-likelihood   n df       BIC       ICL
##       -593.6292 126  6 -1216.276 -1333.505
## 
## Clustering table:
##  1  2  3 
## 15 95 16 
## 
## Mixing probabilities:
##         1         2         3 
## 0.3371770 0.5253839 0.1374390 
## 
## Means:
##         1         2         3 
##  1.858188 10.425551 68.603082 
## 
## Variances:
##        1        2        3 
## 384.2556 384.2556 384.2556

According to the Bayesian information criterion, a model with two clusters describes the data best.

Fitted model:

aucs_diff %>% 
  ggplot(aes(x=diff_of_diffs)) + 
    geom_jitter(aes(y=0, color=first_speaker_type.x), width=0, height=0.001)  +
    ggtitle("Raw data + Components of gaussian mixture") + 
    stat_function(fun = dnorm, args = list(mean = fit2$parameters$mean[1], sd = sqrt(fit2$parameters$variance$sigmasq[1]))) + 
    stat_function(fun = dnorm, args = list(mean = fit2$parameters$mean[2], sd = sqrt(fit2$parameters$variance$sigmasq[2])))

## Warning: Removed 101 row(s) containing missing values (geom_path).

Compute likelihoods based on the adaptation model

## Generalized linear mixed model fit by maximum likelihood (Laplace
##   Approximation) [glmerMod]
##  Family: binomial  ( logit )
## Formula: most_likely_model ~ condition + test_order + first_speaker_type +  
##     first_speaker_type * condition + (1 | workerid)
##    Data: d.post_test
## 
##      AIC      BIC   logLik deviance df.resid 
##    298.6    319.7   -143.3    286.6      240 
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.8870 -0.6145 -0.2432  0.6899  2.7631 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  workerid (Intercept) 1.693    1.301   
## Number of obs: 246, groups:  workerid, 123
## 
## Fixed effects:
##                                              Estimate Std. Error z value
## (Intercept)                                  -0.47550    0.21483  -2.213
## conditioncautious                            -1.00820    0.21698  -4.647
## test_orderparallel                           -0.44470    0.21172  -2.100
## first_speaker_typecautious                    0.61038    0.22015   2.773
## conditioncautious:first_speaker_typecautious  0.08209    0.16821   0.488
##                                              Pr(>|z|)    
## (Intercept)                                   0.02687 *  
## conditioncautious                            3.38e-06 ***
## test_orderparallel                            0.03569 *  
## first_speaker_typecautious                    0.00556 ** 
## conditioncautious:first_speaker_typecautious  0.62551    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtnc tst_rd frst__
## conditincts  0.264                     
## tst_rdrprll  0.131  0.239              
## frst_spkr_t -0.193 -0.318 -0.138       
## cndtncts:__ -0.125 -0.115 -0.028  0.106

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: likelihood_ratio ~ condition + test_order + first_speaker_type +  
##     first_speaker_type * condition + (1 | workerid)
##    Data: d.post_test
## 
## REML criterion at convergence: 3268.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.2301 -0.6277 -0.1118  0.4681  3.1266 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  workerid (Intercept)  7336     85.65  
##  Residual             33961    184.29  
## Number of obs: 246, groups:  workerid, 123
## 
## Fixed effects:
##                                              Estimate Std. Error      df
## (Intercept)                                    -7.145     14.065 120.000
## conditioncautious                             -65.004     11.753 121.000
## test_orderparallel                            -17.405     14.061 120.000
## first_speaker_typecautious                     47.939     14.065 120.000
## conditioncautious:first_speaker_typecautious    3.278     11.753 121.000
##                                              t value Pr(>|t|)    
## (Intercept)                                   -0.508  0.61238    
## conditioncautious                             -5.531 1.87e-07 ***
## test_orderparallel                            -1.238  0.21821    
## first_speaker_typecautious                     3.408  0.00089 ***
## conditioncautious:first_speaker_typecautious   0.279  0.78079    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtnc tst_rd frst__
## conditincts  0.000                     
## tst_rdrprll  0.008  0.000              
## frst_spkr_t -0.024  0.000  0.008       
## cndtncts:__  0.000 -0.024  0.000  0.000

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: likelihood_ratio ~ condition + test_order + first_speaker_type +  
##     prior_likelihood_ratio + first_speaker_type * condition +  
##     (1 | workerid)
##    Data: d.post_test
## 
## REML criterion at convergence: 3266.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.1792 -0.5692 -0.1000  0.4569  3.4660 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  workerid (Intercept)  6588     81.16  
##  Residual             33961    184.29  
## Number of obs: 246, groups:  workerid, 123
## 
## Fixed effects:
##                                              Estimate Std. Error       df
## (Intercept)                                   11.6711    16.2900 119.0000
## conditioncautious                            -65.0040    11.7532 121.0000
## test_orderparallel                           -10.6059    14.1863 119.0000
## first_speaker_typecautious                    46.7980    13.8568 119.0000
## prior_likelihood_ratio                         0.2238     0.1021 119.0000
## conditioncautious:first_speaker_typecautious   3.2781    11.7532 121.0000
##                                              t value Pr(>|t|)    
## (Intercept)                                    0.716  0.47511    
## conditioncautious                             -5.531 1.87e-07 ***
## test_orderparallel                            -0.748  0.45617    
## first_speaker_typecautious                     3.377  0.00099 ***
## prior_likelihood_ratio                         2.193  0.03026 *  
## conditioncautious:first_speaker_typecautious   0.279  0.78079    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtnc tst_rd frst__ prr_l_
## conditincts  0.000                            
## tst_rdrprll  0.122  0.000                     
## frst_spkr_t -0.040  0.000  0.000              
## prr_lklhd_r  0.527  0.000  0.219 -0.038       
## cndtncts:__  0.000 -0.024  0.000  0.000  0.000

## Data: d.post_test
## Models:
## model1: likelihood_ratio ~ condition + test_order + first_speaker_type + first_speaker_type * condition + (1 | workerid)
## model2: likelihood_ratio ~ condition + test_order + first_speaker_type + prior_likelihood_ratio + first_speaker_type * condition + (1 | workerid)
##        npar    AIC    BIC  logLik deviance  Chisq Df Pr(>Chisq)  
## model1    7 3317.7 3342.3 -1651.9   3303.7                       
## model2    8 3314.9 3342.9 -1649.4   3298.9 4.8726  1    0.02729 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

List of adapters:

workerid	first_speaker_type	test_order	noticed_manipulation	cautious_count	confident_count	aligned_count	first_adaptation_speaker_count
1352	cautious	reverse	1	1	1	2	1
1353	confident	parallel	0	1	1	2	1
1354	cautious	parallel	0	1	1	2	1
1355	confident	reverse	0	1	1	2	1
1359	confident	reverse	0	1	1	2	1
1365	cautious	reverse	0	1	1	2	1
1368	confident	parallel	1	1	1	2	1
1370	confident	parallel	1	1	1	2	1
1373	cautious	reverse	1	1	1	2	1
1375	confident	reverse	1	1	1	2	1
1385	cautious	reverse	1	1	1	2	1
1387	cautious	reverse	1	1	1	2	1
1391	cautious	reverse	1	1	1	2	1
1395	cautious	parallel	1	1	1	2	1
1397	cautious	reverse	1	1	1	2	1
1407	confident	reverse	0	1	1	2	1
1417	cautious	reverse	1	1	1	2	1
1423	cautious	parallel	1	1	1	2	1
1428	cautious	parallel	1	1	1	2	1
1432	confident	reverse	0	1	1	2	1
1507	cautious	reverse	0	1	1	2	1
1513	confident	parallel	1	1	1	2	1
1522	confident	reverse	1	1	1	2	1
1524	confident	reverse	0	1	1	2	1
1525	cautious	reverse	1	1	1	2	1
1527	cautious	parallel	1	1	1	2	1
1528	confident	parallel	1	1	1	2	1
1530	cautious	reverse	1	1	1	2	1
1532	cautious	parallel	0	1	1	2	1
1537	confident	parallel	1	1	1	2	1
1539	cautious	reverse	1	1	1	2	1
1541	cautious	parallel	1	1	1	2	1
1543	confident	reverse	0	1	1	2	1
1545	confident	parallel	0	1	1	2	1
1547	cautious	reverse	1	1	1	2	1
1548	cautious	parallel	1	1	1	2	1
1551	confident	reverse	0	1	1	2	1
1552	confident	parallel	1	1	1	2	1
1553	confident	parallel	0	1	1	2	1
1558	confident	reverse	1	1	1	2	1
1559	confident	reverse	1	1	1	2	1
1560	confident	parallel	1	1	1	2	1
1564	cautious	parallel	0	1	1	2	1
1566	confident	reverse	1	1	1	2	1
1571	cautious	reverse	0	1	1	2	1

List of reverse adapters:

workerid	first_speaker_type	test_order	noticed_manipulation	cautious_count	confident_count	aligned_count	first_adaptation_speaker_count
1404	confident	parallel	1	1	1	0	1
1406	cautious	parallel	0	1	1	0	1
1419	confident	reverse	0	1	1	0	1
1521	cautious	parallel	0	1	1	0	1
1546	cautious	reverse	0	1	1	0	1